OP0218 MORTALITY IN PATIENTS WITH PSORIATIC ARTHRITIS IN SWEDEN

Abstract

Background:In contrast to the increased mortality reported in other inflammatory diseases such as rheumatoid arthritis and psoriasis, prior mortality studies in psoriatic arthritis (PsA) have shown inconsistent results.Objectives:To compare all-cause mortality between PsA patients in Sweden and matched general population controls, and to describe cause of death distributions in the two groups.Methods:All individuals in Sweden with ≥1 main diagnosis of PsA (ICD-10: L40.5/M07.0-M07.3) from outpatient visits to rheumatology or internal medicine clinics at age ≥18 years (y) 2001-2017 were identified from the Swedish National Patient Register. Each case was matched to 5 general population controls based on sex, county and age in the year of the first registered arthritis diagnosis for the case. Cases and controls were followed from 1 Jan, 2007, or from first PsA diagnosis thereafter for index cases, until first occurrence of death (data from the Swedish Cause of Death Register), emigration or 31 Dec, 2018. Mortality was assessed overall, as well as stratified by sex (45% males) and disease duration (PsA diagnosis prior to 2007 [38% of cases] vs. 2007-2017), using matched Cox proportional hazard regression, or – in case the Cox assumption regarding proportionality did not hold – matched Breslow test. To account for potential PsA misclassification (in a previous validation study, 86% of 400 cases fulfilled PsA classification criteria), a sensitivity analysis was performed by randomly replacing 20% of cases with one of their own controls. Moreover, incidence rate ratios (IRR) of death were calculated overall and stratified by sex, disease duration and age. Finally, causes of death (from the Cause of Death Register) were described for PsA cases and controls.Results:Over the 12y follow-up, 3 121 deaths occurred among 33 036 PsA cases (268 402 person-years at risk) and 12 884 deaths among 161 144 controls (1 302 250 person-years), resulting in an increased mortality among the PsA cases (HR 1.11 [95%CI 1.07-1.16], p<0.001, Figure and Table; sensitivity analysis HR 1.09 [1.05-1.14]). The increased mortality was seen mainly among female PsA cases and among cases with longer disease duration (Figure; Table). IRR:s of death were significantly increased for all ages except <40y, with the numerically highest point-estimates for ages 40-49y and 50-59y (Table). Cause of death frequencies among the PsA cases/controls: cardiovascular disease 29/27%; diabetes mellitus 2.1/2.5%; chronic kidney disease 0.4/0.3%; infection 5.7/4.5%; chronic pulmonary disease 5.1/4.1%; malignancy 29/34%; suicide 2.3/2.0%; other 27/26%.Table 1.Mortality rates and incidence rate ratiosPsA casesPopulation controlsNumber of deathsPerson-yearsat riskMortality rate*Number of deathsPerson-yearsat riskMortality rate*Incidence rate ratio (95%CI)Overall3 121268 40211.612 8841 302 2509.91.18 (1.13-1.22)Males1 459120 51712.16 468580 28511.11.09 (1.03-1.15)Females1 662147 88611.26 416721 9668.91.27 (1.20-1.34)Longer disease duration1 943139 37913.97 459670 17411.11.25 (1.19-1.32)Shorter disease duration1 178129 0239.15 425632 0778.61.06 (1.00-1.13)Age intervals (years)<401833 5680.598163 2780.60.89 (0.54-1.48)40-499050 5521.8322246 9551.31.37 (1.08-1.73)50-5928065 8204.31 131321 7303.51.21 (1.06-1.38)60-6972370 22410.33 132341 5879.21.12 (1.04-1.22)70-7996037 23225.84 160178 90923.31.11 (1.03-1.19)≥801 05011 00795.44 04149 79181.21.18 (1.10-1.26)* Per 1000 person-years.Conclusion:In this nationwide 12y assessment, the mortality risk among PsA patients in Sweden was increased by around 10% as compared to the general population, mainly driven by increased risks among females and patients with longer disease duration. Cause of death distributions were numerically similar between PsA cases and controls.